Calculate the TPS of the GRID M60-4A on local AI models

NVIDIA 4 GB GDDR5 160 GB/s August 2015

Every model in our catalogue assessed against this card at the context length and minimum quality you choose. Speed is an estimate for a single request, calculated from this card's memory bandwidth and the size of each model once compressed.

Calculated for this card

105 models it can run

721 models in our catalogue altogether

Largest model it holds

DeciLM 6B

5.7B · Q3_K_M · 27.3 tok/s

Fastest model

Gemma 4 E2B

67.8 tok/s · 5.1B

Which AI models can run on a GRID M60-4A?

Set the inputs, read the answer

More context means more memory for the conversation cache. Speed is for a fresh conversation and does not change with this setting.

Hides models that would only fit by being compressed below this point.

105 models match

Calculating
Quantisation Fit
67.8 tok/s

24–135 · low confidence

Gemma 4 E2B 5.1B Apr 2026 3.4 GB 11k tokens ? Q3_K_M Tight
57.7 tok/s

20–115 · low confidence

Gemma 3 1B 1B Mar 2025 1.8 GB 33k tokens Q8_0 Comfortable
57.7 tok/s

20–115 · low confidence

Gemma 3 QAT 1B 1B Apr 2025 1.8 GB 33k tokens Q8_0 Comfortable
57.7 tok/s

20–115 · low confidence

HGRN 1B (WT 103) 1B Nov 2023 1.8 GB 131k tokens ? Q8_0 Comfortable
57.7 tok/s

20–115 · low confidence

LLama 3..2 Typhoon 2 1B 1B Dec 2024 1.8 GB 131k tokens ? Q8_0 Comfortable
57.7 tok/s

20–115 · low confidence

OLMo-1B 1B Feb 2024 1.8 GB 131k tokens ? Q8_0 Comfortable
57.7 tok/s

20–115 · low confidence

Pythia-1b 1B Apr 2023 1.8 GB 131k tokens ? Q8_0 Comfortable
53.5 tok/s

19–107 · low confidence

OpenELM-1.1B 1.1B May 2024 1.9 GB 131k tokens ? Q8_0 Comfortable
52.5 tok/s

18–105 · low confidence

DeciCoder-1B 1.1B Aug 2023 1.9 GB 131k tokens ? Q8_0 Comfortable
52.5 tok/s

18–105 · low confidence

SantaCoder 1.1B Jan 2023 1.9 GB 131k tokens ? Q8_0 Comfortable
52.5 tok/s

18–105 · low confidence

TinyLlama-1.1B (1T token checkpoint) 1.1B Oct 2023 1.9 GB 131k tokens ? Q8_0 Comfortable
52.5 tok/s

18–105 · low confidence

TinyLlama-1.1B (3T token checkpoint) 1.1B Oct 2023 1.9 GB 131k tokens ? Q8_0 Comfortable
48.1 tok/s

17–96 · low confidence

EXAONE 4.0 (1.2B) 1.2B Jul 2025 2.0 GB 131k tokens ? Q8_0 Comfortable
48.1 tok/s

17–96 · low confidence

LFM2-1.2B 1.2B Jul 2025 2.0 GB 131k tokens ? Q8_0 Comfortable
48.1 tok/s

17–96 · low confidence

MinerU2.5 1.2B Sep 2025 2.0 GB 131k tokens ? Q8_0 Comfortable
48.1 tok/s

17–96 · low confidence

Pleias 1.0 1.2B 1.2B Dec 2024 2.0 GB 131k tokens ? Q8_0 Comfortable
48.1 tok/s

17–96 · low confidence

Pleias-RAG-1B 1.2B Apr 2025 2.0 GB 131k tokens ? Q8_0 Comfortable
47.0 tok/s

16–94 · low confidence

Llama 3.2 1B 1.2B Sep 2024 2.2 GB 54k tokens Q8_0 Comfortable
46.3 tok/s

16–93 · low confidence

MiniCPM-1.2B 1.2B Jun 2024 2.0 GB 131k tokens ? Q8_0 Comfortable
44.4 tok/s

16–89 · low confidence

DeepSeek Coder 1.3B 1.3B Jan 2024 2.1 GB 131k tokens ? Q8_0 Comfortable
44.4 tok/s

16–89 · low confidence

DeepSeek-VL-1.3B 1.3B Mar 2024 2.1 GB 131k tokens ? Q8_0 Comfortable
44.4 tok/s

16–89 · low confidence

DigiRL 1.3B Jun 2024 2.1 GB 131k tokens ? Q8_0 Comfortable
44.4 tok/s

16–89 · low confidence

GLA Transformer 1.3B 1.3B Aug 2024 2.1 GB 131k tokens ? Q8_0 Comfortable
44.4 tok/s

16–89 · low confidence

Janus 1.3B 1.3B Oct 2024 2.1 GB 131k tokens ? Q8_0 Comfortable
44.4 tok/s

16–89 · low confidence

Kosmos-2.5 1.3B Aug 2024 2.1 GB 131k tokens ? Q8_0 Comfortable

Speeds are estimates for a single request — one conversation at a time — calculated from memory bandwidth, model size and quantisation. Real throughput varies with the inference runtime and its version. Figures published by hardware vendors measure many simultaneous requests and are much higher.

On record

GRID M60-4A full specification

Everything on record for this board, ordered by how much it bears on running a language model rather than by how a spec sheet would list it. Memory comes first because it decides the outcome; the rest is context.

Memory

The two specifications that decide what this card can run and how quickly. Capacity sets which models fit; bandwidth sets how many tokens per second they produce once they do.

Memory size
4 GB
Memory bandwidth
160 GB/s
Memory type
GDDR5
Memory bus width
256 bit
Memory clock
1.25 GHz

The chip

Which processor is on the board and how it was manufactured. A smaller process size generally means more performance for the same power.

Graphics processor
GM204
Architecture
Maxwell 2.0
Generation
GRID(Mx)
Foundry
TSMC
Process size
28 nm
Transistors
5.2 billion
Transistor density
13,100 K/mm²
Die size
398 mm²
Package
BGA-1745
Released
30 August 2015

Clock speeds

How fast the processor runs. Worth far less here than on a gaming benchmark: generating text is limited by memory bandwidth, so a higher clock barely moves the result.

Base clock
557 MHz
Boost clock
1.18 GHz

Processing units

What the chip contains. These drive graphics performance and matter mainly for processing a long prompt rather than for producing the answer.

Shading units
2,048
Texture mapping units
128
Render output units
64
Streaming multiprocessors
16
L1 cache
48 KB
L2 cache
2 MB

Theoretical performance

Peak arithmetic rates published for the board. These are ceilings that no real workload reaches, and generating text reaches a small fraction of them because it is limited by memory rather than arithmetic.

Single precision (FP32)
4.8 TFLOPS
Double precision (FP64)
150.8 GFLOPS
Pixel rate
75 GPixel/s
Texture rate
151 GTexel/s

The board

What it takes to physically install and power the card — the practical constraints that decide whether it fits the machine you already own.

Power draw (TDP)
225 W
Suggested power supply
550 W
Power connectors
1x 8-pin
Bus interface
PCIe 3.0 x16
Slot width
Dual-slot
Dimensions
267 mm

Software support

Which graphics and compute interfaces the card supports. CUDA compute capability is the one that bears on inference: below 7.0 there are no tensor cores, and modern inference software falls back to slower code paths.

CUDA compute capability
5.2
DirectX
12.1
OpenGL
4.6
Vulkan
1.4
OpenCL
3.0
Shader model
6.8

Listings

Where to buy a GRID M60-4A

No vendor is currently listing this card. Listings come from vendors who publish them here directly — browse the vendor directory to see who is selling what.

What the numbers mean

Memory: the specification that decides everything

Memory

4 GB

Bandwidth

160 GB/s

Largest model

DeciLM 6B

GRID M60-4A carries only 4 GB of GDDR5. That limits it to the smaller end of the catalogue, and a model has to fit entirely inside before it generates anything at all. A runtime actually gets about 3.6 GB.

Memory bandwidth reaches 160 GB/s across a bus of 256 bits. Bandwidth is this card's real constraint. Every token requires reading the entire model out of memory, so a large model will feel slow here even when it fits.

The figure is the bus width multiplied by a memory clock of 1.25 GHz. Both halves matter, and neither is visible in a gaming benchmark.

Put together, the largest model that fits is DeciLM 6B, 5.7B, compressed to Q3_K_M and generating around 27.3 tokens per second.

The chip and how it was built

GRID M60-4A is built on the graphics processor GM204, using the architecture Maxwell 2.0 from NVIDIA, as part of the generation GRID(Mx).

The chip is manufactured by TSMC, on a process of 28 nm, with a die measuring 398 mm², holding 5.2 billion transistors. A smaller process generally means more performance for the same power, though for language models it matters far less than the memory subsystem.

It was released in August 2015, roughly 11.040245647217 years ago. Inference software support tends to follow hardware by a year or two, so a card of this age generally has mature, well-optimised code paths available to it.

Compute throughput, and why it matters less than it looks

FP64

150.8 GFLOPS

Double-precision throughput reaches 150.8 GFLOPS. It has no bearing on running a language model — no inference runtime uses it — but it separates datacentre parts from consumer ones, since the latter deliberately restrict it.

Clocks run from a base of 557 MHz to a boost of 1.18 GHz. Worth far less here than on a gaming benchmark: raising the clock speeds up the arithmetic, and the arithmetic is not what generation is waiting on.

Cache and processing units

GRID M60-4A has an L1 cache of 48 KB, backed by an L2 cache of 2 MB. Cache absorbs a share of the memory traffic that would otherwise hit the main bus, which is the one place on this page where a number other than bandwidth quietly affects generation speed — a large L2 lets more of the working set stay close to the cores.

There are 2,048 shading units, 128 texture mapping units, and 64 render output units. These drive graphics workloads and contribute to prompt processing, but they sit idle for much of the time a model spends generating a reply.

Power, size and installation

Power draw

225 W

GRID M60-4A is rated at 225 W, and the suggested system power supply is 550 W. Running a language model keeps a card busy in bursts rather than continuously — it draws hard while generating and idles between requests — so sustained draw over a working day is usually well below the rated figure.

The board occupies dual-slot, measuring 267 mm long, and needs 1x 8-pin. Worth checking against the case and power supply already in the machine, since the largest cards need considerably more of both than a typical desktop provides.

It connects over PCIe 3.0 x16. The interface governs how quickly a model is loaded from disk into the card, not how fast it runs once there, so a narrower link costs a few seconds at startup and nothing thereafter.

The extremes

The largest AI models that run on a GRID M60-4A

The biggest open-weight models that fit on this card, newest first. Each is shown at the best compression the card can hold.

  1. 01 Gemma 4 E2B 5.1B · Q3_K_M · Apr 2026 67.8 tok/s
  2. 02 Qwen3.5-4B 4B · Q5_K_M · Feb 2026 25.8 tok/s
  3. 03 Nemotron 3 Nano-4B 4B · Q5_K_M · Dec 2025 25.8 tok/s
  4. 04 Qwen3-VL-4B 4B · Q5_K_M · Oct 2025 25.8 tok/s
  5. 05 Qwen3-4B-Thinking-2507 4B · Q5_K_M · Aug 2025 25.8 tok/s
  6. 06 Voxtral Mini 4.7B · Q4_K_M · Jul 2025 28.4 tok/s
  7. 07 Phi-4-Multimodal 5.6B · Q3_K_M · Mar 2025 27.8 tok/s
  8. 08 Minitron 4B 4.2B · Q4_K_M · Nov 2024 31.7 tok/s
  9. 09 XVERSE-MoE-A4.2B 4.2B · Q4_K_M · Apr 2024 31.7 tok/s
  10. 10 DeciLM 6B 5.7B · Q3_K_M · Sep 2023 27.3 tok/s

The fastest AI models on a GRID M60-4A

Where this card produces tokens quickest. Smaller models dominate here, because generating each token means reading the whole model out of memory once.

  1. 01 Gemma 4 E2B 5.1B · Q3_K_M · 3.4 GB 67.8 tok/s
  2. 02 Gemma 3 QAT 1B 1B · Q8_0 · 1.8 GB 57.7 tok/s
  3. 03 Gemma 3 1B 1B · Q8_0 · 1.8 GB 57.7 tok/s
  4. 04 LLama 3..2 Typhoon 2 1B 1B · Q8_0 · 1.8 GB 57.7 tok/s
  5. 05 OLMo-1B 1B · Q8_0 · 1.8 GB 57.7 tok/s
  6. 06 HGRN 1B (WT 103) 1B · Q8_0 · 1.8 GB 57.7 tok/s
  7. 07 Pythia-1b 1B · Q8_0 · 1.8 GB 57.7 tok/s
  8. 08 OpenELM-1.1B 1.1B · Q8_0 · 1.9 GB 53.5 tok/s
  9. 09 TinyLlama-1.1B (1T token checkpoint) 1.1B · Q8_0 · 1.9 GB 52.5 tok/s
  10. 10 TinyLlama-1.1B (3T token checkpoint) 1.1B · Q8_0 · 1.9 GB 52.5 tok/s

Step by step

How to work out the tokens per second of a GRID M60-4A

You do not have to calculate anything by hand — the gputps.com calculator on this page has already worked it out for every model this card can hold. Reading off the answer takes six steps.

  1. 01

    Find the model in the table

    The table lists 105 models this card can run. Search by name, or by size — typing 27b matches on the parameter count even when the name never states it.

  2. 02

    Decide how long your conversations run

    Set the context to your real working length. Short questions cost almost nothing, but a long document can consume a large share of 4 GB so the setting is worth getting right.

  3. 03

    Set a minimum quality if you need one

    Compression is what lets bigger models fit. The quality control drops any model that needs more of it than you are willing to give.

  4. 04

    Look at the range, not just the number

    The figures are calculated, not measured. The fastest result on this card is 67.8 tok/s on Gemma 4 E2B. The same card and model vary by thirty to fifty per cent between inference runtimes.

  5. 05

    Read the fit verdict last

    A tight fit runs but leaves no room to raise the context later; comfortable has headroom. The memory column shows what each model needs against an available 4 GB.

  6. 06

    Open the model to compare cards

    Every model name in the table links to its own page, which runs the same calculation across every card we hold. That is where you see whether the right buy is GRID M60-4A.

Answers

GRID M60-4A — common questions

01

GRID M60-4A— is it good for running local AI models?

Its memory limits it to smaller models though its bandwidth means generation will feel slow on larger models. In total it runs 105 of the models we track. Whether that is enough depends entirely on which model you want — the table above answers that directly.

02

GRID M60-4A— can it run a model that does not fit in its memory?

Only partly. Layers beyond the card's 4 GB sit in system memory and run at a fraction of the speed, so a mostly-offloaded model is rarely worth using. Every figure here assumes it is fully resident on the card.

03

Would two GRID M60-4A cards be twice as fast?

No. A second card doubles the memory to 8 GB of combined memory, at roughly the same generation speed as one.

04

GRID M60-4A— which AI models can it run?

105 of the 721 open-weight language models we track fit on this card and can be run locally. The table on this page lists every one, with the memory it needs, the quantisation it runs at and an estimated generation speed.

05

GRID M60-4A— what is the largest AI model it can run?

The largest model in our catalogue that fits is DeciLM 6B at 5.7B parameters, compressed to Q3_K_M. It generates roughly 27.3 tokens per second and needs about 3.5 GB of the card's memory.

06

GRID M60-4A— how many tokens per second does it produce?

It depends on the model. The fastest model we track here is Gemma 4 E2B at about 67.8 tokens per second, while larger models run proportionally slower because each token requires reading the whole model out of memory once. Speeds are estimates for a single conversation at a time.

07

GRID M60-4A— how much memory does it have?

This card has 4 GB of GDDR5. Around a tenth is reserved by the inference runtime and the driver, leaving roughly 3.6 GB available for a model and its conversation.

08

GRID M60-4A— what is its memory bandwidth?

Memory bandwidth reaches 160 GB/s across a bus of 256 bits. This is the single best predictor of how fast it generates text, because producing each token means reading the entire model out of memory once.

09

GRID M60-4A— what type of memory does it use?

It uses GDDR5 clocked at 1.25 GHz. HBM types are found on datacentre accelerators and carry far more bandwidth than the GDDR used on desktop cards, which is why they generate tokens considerably faster at the same capacity.

10

GRID M60-4A— who makes it?

This is a product of NVIDIA, with the chip manufactured by TSMC, on a process of 28 nm.

11

GRID M60-4A— when was it released?

It was released in August 2015.

12

GRID M60-4A— how much power does it use?

Rated board power is 225 W, and the suggested system power supply is 550 W. Generating text draws hard in bursts and idles between requests, so average consumption over a working session is normally well below the rated figure.

13

GRID M60-4A— how much cache does it have?

The L1 cache is 48 KB, and the L2 cache is 2 MB. Cache absorbs part of the memory traffic that would otherwise reach the main bus, so a larger L2 gives a modest lift to generation speed beyond what bandwidth alone predicts.

14

GRID M60-4A— does it support CUDA?

Yes. It reports CUDA compute capability 5.2, which predates tensor cores. Capability 7.0 and above has tensor cores, which modern inference software uses; below that it falls back to slower code paths for quantised models.

15

GRID M60-4A— what bus interface does it use?

It uses PCIe 3.0 x16. This governs how fast a model is loaded onto the card rather than how fast it runs once loaded, so it costs a few seconds at startup and nothing during generation.

The other direction

Looking at it from the other side?

This page starts from the hardware. If you already know which model you want and need to know what it takes to run it, start from the model instead.

All GPUs